Plumbagin-Serum Albumin Interaction: Spectral, Electrochemical, Structure-Binding Analysis, Antiproliferative and Cell Signaling Aspects with Implications for Anticancer Therapy

Article abstract of DOI:10.1002/cmdc.202000157

Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone) is a small molecule with potent anticancer activity. Like other 1,4-naphthoquinones, it exhibits electrophilic reactivity towards biological nucleophiles. We demonstrate that plumbagin and structurally related 1,4-naphthoquinones with at least one unsubstituted quinoid carbon (C2 or C3) bind to albumin, an ubiquitously present nucleophile, with minimum recovery of free drug. Extraction recovery of plumbagin from albumin in solution showed one-phase exponential decline with a half-live of 9.3 min at 10 μmol/L. In the presence of albumin, plumbagin exhibited instant changes in UV/Vis absorption bands. Electrochemical analysis using cyclic voltammetry showed a decrease in redox peak currents over time until electro-inactivity, thus suggesting the formation of a supramolecular adduct inaccessible for electron transfer. The adduct inhibited cell growth and caused cell-cycle arrest of prostate cancer cells, in part by decreasing levels of the cell-cycle regulator RBBP. The conjugate displayed similar cellular effects to those described for plumbagin, such as decreased levels of androgen receptor and protein kinase C epsilon. The effect of plumbagin-albumin on cancer cells was species-specific, suggesting a receptor-mediated mechanism. Furthermore, it was blocked by cathepsin inhibitor pepstatin A, indicating that lysosomal degradation is involved in the processing of plumbagin-albumin adduct. The spontaneously formed adduct of plumbagin with serum albumin is likely to mediate the biological activities of plumbagin in vivo and to fundamentally influence its pharmacodynamics.

Full text of DOI:10.1002/cmdc.202000157

Accepted Article
Title: Plumbagin-Serum Albumin Interaction: Spectral, Electrochemical,
Structure-Binding Analysis, Anti-proliferative and Cell Signaling
Aspects with Implications for Anticancer Therapy
Authors: Veronique Baron, Adrian Chrastina, John Welsh, gaelle
rondeau, parisa abedinpour, and per borgstrom
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.202000157
Link to VoR: https://doi.org/10.1002/cmdc.202000157
0
1/2020

ChemMedChem
10.1002/cmdc.202000157
Plumbagin-Serum Albumin Interaction: Spectral, Electrochemical, Structure-
Binding Analysis, Anti-proliferative and Cell Signaling Aspects with Implications
for Anticancer Therapy
2§
1
1
2
1
Adrian Chrastina , John Welsh , Gaelle Rondeau , Parisa Abedinpour , Per Borgström and
Véronique T. Baron 1§
1Vaccine Research Institute of San Diego, San Diego Science Center, San Diego, CA 92109
2Proteogenomics Research Institute for Systems Medicine, La Jolla, CA 92037
§Corresponding authors:
A. Chrastina: current address: Proteogenomics Research Institute for Systems Medicine (PRISM), 505
Coast Blvd. South, La Jolla, CA 92037
V.T. Baron: Vaccine Research Institute of San Diego (VRISD), San Diego Science Center, 3030
Bunker Hill, San Diego, CA 92109, USA. Phone: (858) 775-1736; Email: vbaron@sdibr.org
Running title: Plumbagin-albumin interaction and anticancer activity of plumbagin
Keywords: Prostate cancer, plumbagin, albumin
Disclosures: This work was supported by funds from Pellficure Pharmaceuticals, Inc.
P. Borgström, A. Chrastina, P. Abedinpour and V. Baron are inventors on patent(s) in relation to this
work. P. Borgström is President of PellFiCure Pharmaceuticals, Inc. while V. Baron, P. Abedinpour
and A. Chrastina are current or former consultants for the company. J. Welsh and G. Rondeau declare
no competing interest.
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
ABSTRACT
Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone) is a small molecule with potent anticancer
activity. Like other 1,4-naphthoquinones, it exhibits electrophilic reactivity towards biological
nucleophiles. We demonstrate that plumbagin and structurally related 1,4-naphthoquinones with at
least one unsubstituted quinoid carbon (C2 or C3) bind to albumin, an ubiquitously present
nucleophile, with minimum recovery of free drug. Recovery of plumbagin from albumin declined in a
one-phase kinetic with a half-live of 9.3min at 10μM. Plumbagin in the presence of albumin exhibited
instant changes in UV-VIS absorption bands. Electrochemical analysis using cyclic voltammetry
showed a decrease in redox peak currents over time until electro-inactivity, suggesting formation of a
supramolecular adduct inaccessible to electron transfer. The adduct inhibited cell growth and caused
cell cycle arrest of prostate cancer cells, in part by decreasing levels of cell cycle regulator RBBP. The
conjugate displayed similar cellular effects as described for plumbagin such as decreased levels of
androgen receptor and protein kinase C-epsilon. The effect of plumbagin-albumin on cancer cells was
species-specific, suggesting a receptor-mediated mechanism. It was blocked by cathepsin inhibitor
pepstatin A, indicating that lysosomal degradation releases plumbagin to the cell. The spontaneous
formation of a plumbagin-albumin adduct is likely to have crucial implications for the
pharmacodynamics of plumbagin.
-
2 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
INTRODUCTION
Plumbagin (Figure 1A) is a naturally occurring 1,4-naphthoquinone and analog of vitamin K3 isolated
from plants of the Plumbaginaceae, Droseraceae, and Ebenaceae families [1, 2]. Purified plumbagin
inhibits the proliferation of various types of cancer cells, as reviewed extensively in [3, 4], therefore it
is considered a valid oncology drug candidate. In vivo studies have shown that it reduces primary
tumor burden across cancer types as measured by tumor growth rate as well as tumor weight and
volume [4]. Plumbagin inhibits the growth of metastatic foci and delays or suppresses metastatic
colonization in mouse models of breast cancer [5], prostate cancer [6] and lung cancer [7]. In addition,
the drug appears to improve the efficacy of radiation therapy and of various modalities of
chemotherapy in several types of cancer [4]. If these effects are validated, plumbagin may become a
useful adjunct to current standard therapy in a variety of settings. In prostate cancer for example,
plumbagin was most effective in combination with androgen deprivation therapy, which is the clinical
standard of care for hormone-dependent prostate cancer [8, 9]. Plumbagin has now entered clinical trial
for the treatment of prostate cancer in combination with androgen deprivation therapy [10].
The pharmacodynamic properties of plumbagin are fundamental for its efficacy in the clinic.
Pharmacokinetic analyses indicate that the plasma levels of plumbagin decrease sharply with time in
animals, particularly when administered intravenously [11, 12]. The quinoid pharmacophore of
plumbagin and other 1,4-naphthoquinones exhibit potent electrophilic reactivity towards biological
nucleophiles. The high reactivity of plumbagin added to its rapid plasma clearance suggest that it may
rapidly interact with plasma nucleophiles such as serum albumin, which is the most abundant protein
in the plasma, contributing to about 60% of the blood proteins (35–50 g/L in human serum). Indeed,
structurally related naphthoquinone derivatives bind to bovine serum albumin (BSA) as well as human
serum albumin (HSA) in vitro [13]. Serum albumin plays a critical role in the maintenance of blood
colloidal pressure and acts as a solubilizer and detoxifying protein. It represents the main transport
carrier in the blood circulation for fatty acids, ions such as calcium, steroid hormones and many
systemically administered pharmaceutical drugs, including antibiotics, anticoagulants, anti-
inflammatory drugs, anesthetics, benzodiazepines, etc [14, 15]. Binding of drugs to albumin alters their
metabolism, distribution and physiological effects [16, 17]. It increases the solubility of an otherwise
poorly soluble drugs and slows its distribution through the tissues while decreasing its side-effects. It
also improves the pharmacokinetic profile of the drug since serum albumin has a long half-life of about
1
9 days owing to its interaction with the Fc receptor [18]. Importantly, serum albumin can potentially
-
3 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
act as an effective carrier for drug delivery to tumors. Several studies in animals showed that albumin
is accumulated in tumor tissue due to increased vascular permeability and lack of functional lymphatic
system, a mechanism known as enhanced permeability and retention (EPR) [19, 20]. Furthermore,
there are several albumin-binding receptors, including gp60 [21], SPARC [14, 22], and scavenger
receptors gp18 and gp30 [23] that mediate specific interaction with serum albumin and can facilitate
the selective uptake of albumin in tumor tissue.
The present study analyzes the spontaneous interaction of plumbagin with albumin, leading to the
formation of a plumbagin-albumin adduct that exhibits anti-proliferative activity against cancer cells.
The plumbagin-albumin adduct increased the phosphorylation of AMPK and decreased protein levels
of PKC epsilon and AR in prostate cancer cells, which are known effects of plumbagin. The
extemporaneous interaction of plumbagin with albumin, a ubiquitously present nucleophile, could have
a fundamental influence on the pharmacokinetics and the anticancer activity of plumbagin.
-
4 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
MATERIALS AND METHODS
The 2,3-dimethyl-1,4-naphthoquinone was purchased from SIA MolPort (Latvia, EU). Plumbagin,
juglone, naphthazarin, 1,4-naphthoquinone, menadione and other common chemicals were from
Sigma-Aldrich (Saint-Louis, MO, USA). Peroxidase-conjugated F(ab’)2 antibodies were from
Rockland (Limerick, PA, USA) and the cell cycle sampler kit (611423) was from BD Biosciences (San
Jose, CA, USA). The PKC Isoform Antibody Sampler Kit (#9960), pan-phospho Ser660-PKC (#9371);
AMPK and ACC Antibody Sampler Kit (#9957), anti-PSA antibody (#2475) were purchased from Cell
Signaling Technology (Danvers, MA, USA). Peroxidase-conjugated antibodies against AR (sc-815)
and against actin (sc-47778) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA).
Synthesis of 2,3-dimethoxy-1,4-naphthoquinone (DMNQ)
0
.01 mol of 2,3-dichloro-1,4-naphthoquinone and 0.03 mol of sodium methoxide were refluxed in 50
ml of anhydrous methanol for 4h. At that time 0.02 mol of sodium methoxide was added to the
reaction and the mixture was refluxed for 1h. The reaction product was concentrated under vacuum
and the solid residue was filtered off and extensively washed with water. Yellow solid, UV-VIS (MeOH)
3
−1
−1
1

max (nm) (εmax, dm .mol .cm ): 247 (20,200), 276 (13,700), 330 (2,800). H NMR (500 MHz, DMSO-d ), δ,
6
ppm: 7.96 (dd, J = 5.7, 3.3 Hz, 2H), 7.84 – 7.79 (m, 2H), 3.99 (s, 6H). ESI-MS, m/z: 241.1 [M+Na]+.
Synthesis of 5-O-acetylplumbagin (6-methyl-5,8-dioxo-5,8-dihydronaphthalen-1-yl acetate)
o
1
mmol of plumbagin in 10 ml of dichloromethane was mixed with 3 mmol of pyridine at 0 C and 2
o
mmol of acetyl chloride was added while mixture was stirred at 0 C. The reaction mixture was
incubated for 4 h at room temperature then washed with water and brine. The dried organic phase was
o
then resolved by column chromatography on silica to yield product. Yellow solid, mp 118-120 C. UV-
3
−1
−1
1
VIS (MeOH) max (nm) (εmax, dm .mol .cm ): 245 (13,800), 252 (13,500), 337 (2,600). H NMR (500 MHz,
DMSO-d ), δ, ppm: 7.97 (dd, J = 7.7, 1.3 Hz, 1H), 7.88 (t, J = 7.9 Hz, 1H), 7.56 (dd, J = 8.0, 1.3 Hz, 1H), 6.85
q, J = 1.5 Hz, 1H), 2.35 (s, 3H), 2.09 (s, 3H). ESI-MS, m/z: 253.2 [M+Na]+.
6
(
Formation of the plumbagin-albumin conjugate
Plumbagin was incubated with bovine serum albumin (BSA) or human serum albumin (HSA) in
equimolar ratio in phosphate-buffered saline (PBS) pH7.4, for 18h at 37ºC. The concentration of bound
and free plumbagin was determined UV-VIS spectrophotometry (DU-640, Beckamn Coulter) after
chloroform extraction. The conjugate was dialyzed against PBS for 24h (Spectra/Por 12-14kDa
MWCO, Spectrumlabs) and then concentrated by membrane ultrafiltration (Millipore MWCO10K).
-
5 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
The final concentration of albumin was determined using a BCA kit according to manufacturer's
instructions (Pierce).
Cyclic voltammetry
Cyclic voltammetry measurements were performed using a computer-controlled Autolab potentiostat
PGSTAT-101 (Metrohm Autolab B. V., Netherlands) and NOVA software (Metrohm Autolab B. V.).
All experiments were conducted at 25℃ in a conventional tree-electrode cell that consisted of a glassy
carbon working electrode (diameter 3 mm), a platinum wire as auxiliary electrode, and an
Ag/AgCl/KClsat reference electrode. The surface of the glassy carbon electrode was polished with 0.3
µm alumina powder and extensively rinsed with deionized water to obtain clean electrode surface
before each measurement. Before all measurements, the cell was purged with high-purity argon.
Measurements were performed at scan rate 100 mV/s.
Recovery of plumbagin, warfarin and plumbagin analogues from serum albumin
Plumbagin (10 and 100 µmol/L), warfarin (100 µmol/L), or various 1,4-naphthoquinones (100 µmol/L:
juglone, naphthazarin, 1,4-naphthoquionone, menadione, 5-O-acetylplumbagin, 2,3-dimethyl-1,4-
naphthoquionone and 2,3-dimethoxy-1,4-naphthoquionone), were incubated with BSA (45 mg/ml in
PBS, pH7.4) at 37ºC. Four replicate samples were collected from each mixture of drug/albumin at
designed time-points and used for extraction. Briefly, 0.4 ml of sample was extracted with equal
volume of chloroform by constant mixing at room temperature for 2 min using a vortex. Then, equal
volume of trichloroacetic acid (20% v/v) was added to each sample and the mixture was mixed for 2
min. Samples were centrifuged for 5 min at 12,000 rpm. 0.1 ml aliquots of chloroform layer were
evaporated. Solid residues were dissolved in 0.2 ml of methanol. The concentration of a particular drug
was determined using UV-VIS spectrophotometry (DU-640, Beckamn Coulter) at its respective
3
−1
−1
absorption peak; UV-VIS (MeOH) max (nm) (εmax, dm .mol .cm ): plumbagin, 262 (11,800); warfarin,
04 (8,750); juglone, 246 (10,900); naphthazarin, 267 (5,080); 1,4-naphthoquionone, 242 (18,170); 5-
O-acetylplumbagin, 245 (13,800); 2,3-dimethyl-1,4-naphthoquionone, 329 (14,700); 2,3-dimethoxy-
,4-naphthoquionone, 247 (20,200)). Data were normalized to initial concentration and expressed as
3
1
means with corresponding Standard Deviations.
Cells
PTEN-P2 mouse prostate cancer cells generously provided by Dr. Wu [24] were grown in phenol red-
free DMEM medium containing 10% heat-inactivated fetal bovine serum (FBS), 2 mmol/L L-
-
6 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
-8
glutamine, 100 U/ml penicillin/100 μg/ml streptomycin and DHT 10 mol/L final. Human cancer cells
LNCaP (prostate), DU145 (prostate) and UM-UC-3 (bladder) were purchased from ATCC (American
Tissue Culture Collection, Manassas, VA, USA). LNCaP cells were grown in phenol-free RPMI-1640
-8
medium containing 10% FBS, 4.5 g/l glucose and DHT 10 mol/L final. DU145 and UM-UC-3 cells
were grown in DMEM containing 10% FBS, 2 mmol/L L-glutamine and 100 U/ml penicillin/100
μg/ml streptomycin. Cells were maintained in a humidified incubator at 37ºC and 5% CO .
2
Cell line integrity
To prevent cell line derivation, cells were cultured for a maximum of 20 passages and started over
periodically from vials frozen at very early passages. All cell lines were preemptively treated with an
antimycoplasma reagent (MP Bio, Santa Ana, CA) upon arrival and at each start from a frozen vial. A
mycoplasma-detection assay (Lonza, Allendale, NJ) performed at the end of the study indicated that
the cells were free of mycoplasma. AR-positive prostate cancer cells PTEN-P2 and LNCaP cells have
distinctive characteristics that are verified periodically (response to androgen and AR expression;
LNCaP cells are also PTEN-negative).
Cell growth/viability assay
Cells were plated at a density of 5,000 cells/well in 96-well plates in normal growth medium the day
before treatment. Increasing concentrations of plumbagin-albumin conjugate, up to 20 µmol/L
(plumbagin equivalent) were added to the cells for 24hrs. A mass equivalent concentration of serum
albumin was used as control. Cell numbers were assessed by adding 10 µL of WST-1 reagent (water
soluble tetrazolium assay (2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium;
Takara). After 90 min the optical density (OD) was measured at 430nm in a spectrophotometer
(Molecular Devices; Spectra Max 250). The blank was measured from wells that did not contain cells
and subtracted from all readings.
To assess the effect of plumbagin-albumin on cell growth over time, PTEN-P2 cells were treated with
plumbagin-BSA and cell numbers/viability were determined daily over a period of 6 days using the
WST1 assay. The culture medium was changed every other day. In one set of conditions plumbagin-
albumin was added only at Day 0 (acute treatment), whereas in another set of conditions BSA control
or Plumbagin-BSA were added every day (chronic treatment).
Cell cycle analysis
-
7 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
PTEN-P2 cells were treated with plumbagin-BSA (10 µmol/L) or with control unconjugated BSA
equivalent mass concentration) for 24, 48 and 72 hours. Attached cells were suspended with trypsin
(
and pooled with floating cells potentially present in the medium. After two washes in PBS, cells were
fixed by adding ice-cold ethanol dropwise to reach a final concentration of 70% ethanol and were
maintained at 4C for a minimum of 2hrs. Fixed cells were washed twice with PBS and suspended in
PBS containing 0.1% Triton-X100 (v/v), 50 μg/ml Propidium Iodide and 50 μg/ml DNase-free RNase,
for 30 min at 22ºC. Fluorescence of single cells was measured by Flow Cytometry using the 488 nm
laser on an Accuri C6 instrument (BD Biosciences, Franklin Lakes, NJ). De NovoTM FCS express
Software was used for data analysis. Of note, very few apoptotic cells were seen in the pre-G1 fraction
which contains fragmented DNA and were not included in the analysis.
Western blots
Cells were lysed on ice in RIPA buffer containing phosphatase and protease inhibitors as described
[
25]. Lysates were clarified by centrifugation for 15min at 13,000rpm, protein concentration was
determined using the BCA assay (Pierce, Rockford, IL), and lysates were resuspended in SDS-PAGE
buffer. After SDS-PAGE electrophoresis, proteins were transferred to Immobilon-P® membranes
(Millipore, Billerica, MA). Membranes were incubated with a 5% bovine serum albumin blocking
buffer for 30 min and the first antibody was incubated overnight at 4ºC. Peroxidase-conjugated
antibodies (Amersham Biosciences, Piscataway, NJ) were added for 45 min at 22ºC followed by a 5
min incubation in Western blot Luminata™ HRP substrate (Millipore). Analysis and quantification
were performed using a FluoChem™ instrument 8900 (AlphaInnotech/Protein Simple, Santa Clara,
CA) Membranes were stripped using RestoreTM Stripping Buffer (Pierce/Thermo Scientific, Pittsburg,
PA) for 30 min at 22ºC, then reprobed with the indicated antibodies. Where indicated, results were
quantified using the instrument integrated quantification software (AlphaEase FC).
-
8 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
RESULTS
An analysis of the cell culture parameters that influence the anti-proliferative action of plumbagin
revealed that the presence of serum in the medium had a significant impact on the effect of the drug, as
shown in Supplementary Figure S1. Furthermore, pre-incubation of plumbagin with serum before
exposure to cells altered the pattern of anti-proliferative activity of the drug, indicating another mode
of action of plumbagin when pre-exposed to serum components (Supplementary Figure S2). This
correlates with the observation that plumbagin could not be recovered after 1h of incubation with
serum, indicating formation of adducts with a constituent of the serum (Supplementary Figure S3). The
high reactivity of the drug towards nucleophiles, the observation that it is rapidly sequestered from the
serum, together with its fast clearance from the blood in vivo all suggest that plumbagin reacts with
nucleophiles in the serum such as albumin. This prompted us to analyze the interaction of plumbagin
and its analogs with albumin and to test the hypothesis that plumbagin exerts anti-cancer effects
through a plumbagin-albumin adduct.
Spectral and electrochemical changes induced by interaction of plumbagin with serum albumin
UV-VIS absorption spectroscopy was used to monitor spectral changes following the addition of
plumbagin to bovine serum albumin. The reference light absorption spectra of equivalent
concentrations of plumbagin, DMNQ and BSA are shown in Figure 1C. As shown in Figure 1Da, a
time-dependent decrease of absorption maxima corresponding to the peak absorption of plumbagin
(418 nm) and a concomitant increase of absorption in the near UV range at 350-380 nm demonstrated
the formation of a conjugate between plumbagin and albumin, starting immediately upon mixing with
tendency to plateau at about 3 hours of incubation. On the contrary, there were no spectral changes
observed over time (>30 sec) after initial addition of structural analog of plumbagin, 2,3-dimethoxy-
1
,4-naphthoquinone (DMNQ) to BSA (Figure 1Db). These results indicate that DMNQ in BSA
established a rapid equilibrium of non-covalent interaction while formation of plumbagin-BSA adduct
proceeded over the course of 2-3 hours.
Changes in the electrochemical activity of plumbagin upon addition to albumin are shown in Figure 2.
Electrochemical activity was measured using cyclic voltammetry on a glassy carbon electrode in PBS
as supporting electrolyte. The reference cyclic voltammograms of plumbagin in PBS showed one-peak
single-step redox waves (Figure 2A) with reduction peaks at -0.290 V and oxidation peaks at -0.260 V
(
vs Ag/AgCl at scan rate 100 mV/s), indicating a two-electron redox process that is characteristic of
,4-naphthoquinones in aqueous media [26]. The redox peak currents were proportional to the
9 -
This article is protected by copyright. All rights reserved.
1
-

ChemMedChem
10.1002/cmdc.202000157
concentration of plumbagin (Figure 2B). These characteristics enable us to monitor the
electrochemistry and concentration of free plumbagin in aqueous solution. Electrochemical analysis of
plumbagin upon addition to BSA shows an instant decrease in redox peak currents (Figure 2C).
Gradually over time plumbagin in the BSA solution became electro-inactive, showing a complete
decrease in redox peak currents to the capacitive level (Figure 2D). On the other hand, structural
analog of plumbagin, DMNQ, did not become electrochemically inactive (Figure 2E-F) and after
initial equilibrium (<30 sec) retained redox activity (Figure 2F). Changes in peak anodic current over
time for plumbagin and DMNQ incubated with BSA are shown in Figure 2G. These results point to
the formation of a supramolecular adduct of plumbagin with albumin within which the 1,4-
naphthoquinone moiety of plumbagin is inaccessible for electron transfer.
To further analyze the interaction of plumbagin with albumin we investigated extraction recovery
(amount of free drug that could be recovered from the complex by chloroform extraction) of
plumbagin and several 1,4-naphtoquinone analogs from solutions of albumin (Figure 3). Warfarin, a
drug known to bind non-covalently to albumin, was used to validate the extraction method. Warfarin
instantly bound to albumin as evidenced by the low percentage of free warfarin recovered by
membrane ultrafiltration (Figure 3A). As shown in Figure 3B, 100% of albumin-bound warfarin was
recovered by chloroform extraction at all time points, consistent with the expectation that warfarin
dissociates from the complex upon extraction since binding is reversible. On the other hand, the
amount of free plumbagin that could be extracted from the albumin complex decreased to 0% within 2
hours of incubation, with half-lives of 15.7 min (10 μmol/L) and 18.1 min (100 μmol/L) when
incubated in BSA (Figure 3C), and a half-live of 32.1 min when mixed with HSA (Supplemental
Figure S4), indicative of adduct formation. These results indicate that the interaction between
plumbagin and serum albumin is irreversible.
The structure-binding relationship of plumbagin and some of its analogs was analyzed to identify
functional group(s) of plumbagin involved in adduct formation with albumin. A series of 1,4-
naphthoquinone derivatives (juglone, naphthazarin, 1,4-naphthoquinone, menadione, 2,3-dimethyl-1,4-
naphthoquinone) were acquired from a commercial source, whereas 5-O-acetyl-plumbagin and 2,3-
dimethoxy-1,4-naphthoquinone were synthesized as described in Methods. The chemical structure of
each drug used in these experiments is shown in Figure 3D. Each analog was mixed with albumin and
the amount of analog that could be extracted by the chloroform method was determined by UV-VIS
spectrophotometry. As shown in Figure 3E, Plumbagin, juglone, naphthazarin, 1,4-naphthoquinone,
-
10 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
menadione and 5-O-acetyl-plumbagin were not extractible from the albumin complex (<1%) but
showed almost complete retention after 18 hours of incubation with albumin. On the other hand,
plumbagin analogs 2,3-dimethoxy-1,4-naphthoquinone and 2,3-dimethyl-1,4-naphthoquinone were
completely recovered by extraction, indicating that they did not form an irreversible adduct with
albumin. Thus, 1,4-naphthoquinones with at least one unsubstituted quinoid carbon (C-2 or C-3) can
irreversibly bind to serum albumin, presumably through covalent bond, whereas 1,4-naphthoquinones
with substituted quinoid carbons (both C-2 and C-3) do not form irreversible adducts with serum
albumin. These results indicate the site-specific nucleophilic addition of serum albumin to the C-2 or
C-3 carbon of plumbagin.
Biological effect of plumbagin-albumin
To ascertain the biological activity of the adduct and its potential contribution to the effects of
plumbagin on the growth of tumor cells in vitro, plumbagin-albumin was added to cancer cells and cell
numbers were evaluated 24h later using the WST-1 assay. As shown in Figure 4A, the plumbagin-
BSA adduct caused a dose-dependent growth inhibition in mouse prostate cancer cells PTEN-P2. No
effect of plumbagin-BSA was observed in human prostate cancer cells DU145, whereas plumbagin-
HSA decreased their growth, suggesting that cellular effects of the conjugate are species-specific.
Plumbagin-HSA also inhibited the growth of human bladder cancer cells UM-UC-3 (Figure 4A) and
of human breast cancer cells MCF-7 (data not shown). This species specific effect of the plumbagin-
serum albumin adduct suggests that it is mediated via cell surface receptor-mediated endocytosis.
Various albumin-binding receptors have been identified in human cells including cancer cells
(reviewed in [22]). Since bovine and murine albumin share high sequence homology, plumbagin-BSA
is expected to interact with mouse cells, whereas there is only about 70% homology between bovine
and human albumin, explaining that only the plumbagin-HSA conjugate was active in human cells.
The receptor/albumin-plumbagin complex is then internalized by receptor-mediated endocytosis.
While the receptor is recycled, albumin is subject to lysosomal degradation (reviewed in [27, 28]). To
test whether lysosomal degradation is necessary, we used various protease inhibitors including
pepstatin A, which inhibits lysosomal proteases cathepsins D and E. Thus, serine protease inhibitor
leupeptin, caspase inhibitor Z-VAD, pepsin and cathepsin D/E inhibitor Pepstatin A, and anti-oxidant
Trolox were pre-incubated with cancer cells before addition of plumbagin-BSA or BSA control.
Twenty-four hours later cell proliferation was evaluated by WST1. As shown in Figure 4B,
plumbagin-BSA inhibited cell growth by 50% following a 24h incubation. Leupeptin, Z-VAD and
-
11 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
trolox did not influence this effect, whereas pepstatin A antagonized the effect of plumbagin-albumin,
indicating that lysosomal degradation is necessary for the biological activity of the adduct. Of note,
earlier experiments have shown that pepstatin A does not antagonize the effect of free plumbagin
(Supplementary Figure S5).
The observation that plumbagin-albumin fails to inhibit cell growth in cells pre-treated with pepstatin
A supports the hypothesis that processing of albumin-plumbagin adduct by lysosomal degradation is
necessary for release of the active form of plumbagin. Thus, we hypothesize that following receptor-
mediated internalization and proteolytic degradation of plumbagin-albumin adduct in lysosomes, the
amino acid/peptide-adducts of plumbagin are released into the cytosol, essentially resulting in the post-
catabolic release of a small plumbagin adduct in which plumbagin is accessible, in contrast to the large
intact plumbagin-albumin adduct in which plumbagin is “protected” within the binding pocket. Small
molecule 1,4-naphthoquinone adducts such as the cysteine-plumbagin adduct have been shown to be
redox-active with the same potential range of action as plumbagin [29]. Therefore these very small
adducts are capable of redox cycling to form ROS species and are expected to exert similar anti-
proliferative activity as free plumbagin. On the other hand, digested plumbagin-albumin adducts could
possibly undergo a retro-Michael reaction that would liberate free plumbagin. This will be the subject
of future research aimed at providing more insight into the mechanism of action of the plumbagin-
albumin adduct discovered here.
Plumbagin conjugated with serum albumin did not obviously alter cell morphology, although there
were noticeably fewer cells after plumbagin treatment for 24h than in the control, which is consistent
with an effect on cell growth (Supplementary Figure S6). Of note, the morphology of cells incubated
with free plumbagin was markedly different, with early evidence of catastrophic cell death not visible
in the presence of plumbagin-albumin. In addition, the morphology of cells treated with the adduct was
very similar to conditions in which plumbagin was quenched by serum (Supplementary Figures S1
and S2).
When cell growth was measured over time, it was observed that the adduct completely inhibited cell
growth compared to control-treated cells, as shown in Figure 5A. It is interesting to note that the effect
of an acute one-time treatment with plumbagin-albumin adduct caused growth inhibition for as long as
3
days, suggesting a long half-life of the adduct in vitro. The inhibition was reversible because cells
started growing again 4 days after the one-time treatment. Chronic exposure, on the other hand, led to
-
12 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
prolonged and complete inhibition of cell growth and eventually to some cell death, since the number
of cells at day 6 was half of the number of plated cells (4,939±570 initial versus 2,695±400 at day 6).
To evaluate the effect of plumbagin-albumin adduct on cell cycle progression, non-synchronized
PTEN-P2 cells growing in normal culture medium were treated with either un-conjugated BSA or
plumbagin-BSA for the indicated times. Cells were stained with propidium iodide for flow cytometry
analysis. The proportion of cells in each phase of the cell cycle did not vary significantly as a function
of time when cells were treated with control BSA and a classic pattern of phase distribution was
observed in which phase G1 was predominant and accounted for approximately 50% of the cells
(Figure 5B and C). About 25% of cells were in S phase, as expected from a non-synchronized,
actively growing cell population. It was observed that the proportion of cells in G1 phase decreased
upon treatment with plumbagin-BSA adduct, while the proportion of cells in S phase and G2/M
increased. These results indicate that plumbagin-BSA adduct alters cell cycle progression and are
consistent with a depletion of the cells in G1 and a block in the G2/M phase, either at the G2
checkpoint or in mitosis.
To determine if the effect on the cell cycle could be explained by alterations in the expression or
phosphorylation levels of cell cycle regulators, PTEN-P2 cells were treated with plumbagin-BSA
adduct at increasing concentrations and for variable periods of time. Expression and phosphorylation
levels of cell cycle proteins were analyzed by western blot. Few cell cycle regulators were affected by
plumbagin-BSA adduct. For example, no alteration was observed in cyclin D1, p27KIP1, p21CIP1
,
p19INK4d, CDK1/cdc2, among others (not shown). However, plumbagin-albumin decreased protein
levels of p48-RBBP protein isoforms by ≥ 80% (Figure 5D).
On the other hand, the plumbagin-albumin adduct did not activate tumor suppressor p53 and did not
induce p21 CIP1 in PTEN-P2 cancer cells, which was similar to the lack of effect observed in response
to free plumbagin (Suplementary Figure S7).
We have shown previously that in prostate cancer cells, plumbagin decreases the protein expression of
the androgen receptor (AR) [9] and decreases the hormone-induced expression of AR target genes
[
30]. This is particularly relevant to the effect of plumbagin in prostate cancer because prostate tumors
depend on the AR axis for growth. When AR-positive prostate cancer cells PTEN-P2 were treated with
plumbagin-BSA, a dose-dependent decrease in AR protein levels was observed (Figure 6A). The
effect peaked at 5 µmol/L (Figure 6A) after 16 hours of treatment and lasted at least 24h following
single treatment (Figure 6B). The average AR expression following treatment with 10 µmol/L of
-
13 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
plumbagin conjugate for 24h was 34%±4.5 of control and was statistically significant (p=0.0028, t-test,
Figure 6C). Consistently, exposure to plumbagin-albumin adduct also decreased the expression of AR
target protein PSA (prostate specific antigen) in human prostate cancer cells LNCaP (Figure 6D).
It was also observed that plumbagin-albumin adduct caused a decrease in the protein levels of various
PKC isoforms, including protein levels of PKCε, thereby mimicking the well described effect of
plumbagin on prostate tumors in animals [6, 31-33]. The effect of plumbagin-albumin on PKC levels
was analyzed by treating cells with plumbagin-albumin, followed by western blot analysis. As shown
in Figure 7A, the conjugate decreased the protein expression levels of PKCε, PKCµ and PKCδ with a
maximum effect at 24h. Average PKCε expression following treatment with 10 µmol/L of plumbagin
conjugate for 24h was 36.8% of control (SE±5.0) and was statistically significant (p=0.0026, Figure
7
B). Of note, PKCα and PKCζ were not detectable by western blot in these cells. Figure 7C shows
that treatment with plumbagin-albumin caused a partial decrease in the phosphorylation of PKC
measured using a pan-antibody), which is likely a result of the decreased expression of the
(
corresponding isoform.
Interestingly, the conjugate stimulated the phosphorylation of AMPKα and the phosphorylation of
AMPKα direct target ACC (acetyl-CoA carboxylase). The average AMPKα phosphorylation following
treatment with 10 µmol/L of plumbagin conjugate for 1h was increased 4 fold (p=0.03, Figure 7D).
Because activation of AMPKα is another known effect of free plumbagin [34, 35], these results also
suggest that the conjugate has similar biological effects as unconjugated plumbagin.
We conclude that the plumbagin-albumin adduct mimics at least some of the cellular effects of
plumbagin including known in vivo molecular alterations and therefore is biologically relevant.
-
14 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
DISCUSSION
Serum albumin is the most abundant protein in the plasma (35 - 50 g/L). It binds to a wide range of
drugs, limiting their free concentration and altering their metabolism, distribution and physiological
effects. This can be desirable as it may, for instance, increase the solubility of an otherwise poorly
soluble drug, slow its distribution through the organism, and increase its half-life while decreasing its
side-effects. Albumin therefore has emerged as a potential carrier for therapeutic drugs and several
drug delivery systems that effectively exploits albumin as a drug carrier have entered clinical trials or
are used clinically [15, 36].
Several therapeutics or their reactive metabolites form covalent complexes with serum albumin,
including BI-94 [37] and Neratinib (HKI-272) [38] among others. These typically include soft
electrophiles such as a,ß-unsaturated carbonyl compounds, quinones, quinone imines, quinone
methides, imine methide, isocynate, isothiocynates, aziridinium, etc [39, 40]. The soft electrophiles
react with cysteine residues in proteins acting as soft nucleophilic substrates [39, 41]. Serum albumin
contains a free accessible cysteine residue (Cys-34) [42-44]. The free thiol group of Cys-34 is very
reactive toward electrophiles at physiological pH due to low pKa value of 6.5 [45]. In plasma, free
thiol at Cys-34 of serum albumin (0.5 - 0.8 mM) is the most abundant thiol [46] and as such it is
expected to act as a trap for soft electrophiles.
1
,4-naphothoquinones including plumbagin are highly reactive organic chemical species toward
biological nucleophiles [47]. Added to the observation that the effect of plumbagin is rapidly altered by
serum in cell culture medium, together with its fast clearance from the blood in vivo [11, 12], all
suggest that plumbagin rapidly reacts with endogenous nucleophiles abundant in the serum such as
albumin. Indeed, plumbagin and structurally related 1,4-naphthoquinones with at least one
unsubstituted quinoid carbon (C2 or C3) showed instantaneous formation of supramolecular adduct
with of albumin as supported by extraction recovery experiments. Necessity for the unsubstituted
carbon in 1,4-naphthoquinone core indicates binding via Michael's addition reaction, where plumbagin
and other 1,4-naphthoquinones being α, β-unsaturated diketones act as electrophilic acceptors in 1,4-
adition reaction [47]. These findings support the hypothesis that plumbagin forms covalent adduct with
albumin, most likely through a cysteine residue such as free Cys-34.
Furthermore, it is hypothesized that the adduct specifically interacts with cells via a cell surface
receptor for albumin that is internalized and degraded in the lysosomes, thus releasing the active form
of plumbagin. The plumbagin-albumin adduct profoundly inhibited cell growth and caused cell cycle
-
15 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
arrest. The adduct is biologically relevant and represents a credible intermediary in vivo, as it showed
similar cellular effects as have been described for free plumbagin. For instance, the plumbagin-
albumin adduct inhibited cell growth through cell cycle arrest in the G2/M phase. We observed that
plumbagin-albumin did not activate the p53/p21CIP pathway but reduced the expression of RBBP
family of proteins. RBBP family of proteins are retinoblastoma-binding proteins with histone or DNA-
binding capabilities, functioning as cofactors and regulators of chromatin assembly factors, chromatin
remodeling factors and histone modification enzymes. They play a crucial role in DNA replication and
as cell cycle checkpoints, and therefore in the control of cell proliferation.
We then looked at pertinent cellular responses, the first of which is the well documented decrease in
the expression of PKC-ε in vivo. Thus, plumbagin consistently reduced the expression of PKC-ε in the
tumors of treated animals compared to non-treated animals in various mouse models of prostate cancer
(Aziz et al., 2008; Hafeez et al., 2013; Hafeez et al., 2012). It is noteworthy that different routes of
administration (intra-peritoneal and per oral) tested in various models resulted in similar effects. In
addition, the very consistent decrease of PKC-ε levels in plumbagin-treated tumors points to this
kinase as a crucial target. PKC-ε is a master switch regulator of cell signaling that regulates cancer-
relevant biological functions such as differentiation and proliferation, and therefore may play an
important role in the anti-cancer effect of plumbagin. Its expression is often altered in prostate tumors,
in which it appears to act as an oncogene (reviewed in [48, 49]). Recent studies have also shown that it
may be involved in prostate cancer metastasis [50]. A second target of plumbagin that is especially
pertinent to prostate cancer, for which plumbagin is in clinical trial, is the androgen receptor.
Plumbagin has been shown previously to inhibit AR expression [9] and to decrease the hormone-
induced expression of AR target genes [30]. It is shown here that plumbagin-BSA indeed decreased
AR expression in PTEN-P2 prostate cancer cells. In addition, plumbagin-HSA decreased the AR target
PSA in LNCaP human prostate cancer cells. Prostate tumors depend on AR signaling for growth,
therefore the observation that plumbagin-albumin has a similar inhibitory effect on AR as
unconjugated plumbagin is physiologically relevant. Finally, plumbagin-albumin increased the
phosphorylation of AMPKα and its direct target ACC, a similar effect as described previously for
plumbagin [34, 35]. AMPK is an evolutionarily conserved serine/threonine kinase that functions as a
sensor of cellular energy levels and regulates metabolism. In particular, AMPK stimulates glycolysis
and increases glucose uptake, inhibits glycogen synthesis and overall regulates glucose metabolism. It
also inhibits protein synthesis, activates autophagy, and decreases lipid anabolism while activating
-
16 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
lipid catabolism [51]. Thus, activation of AMPK is known to contribute to the anti-cancer properties of
several drugs such as metformin and is relevant to the anti-cancer effect of plumbagin.
The interaction between serum albumin and drug is therefore an important factor to consider during the
development of a new therapeutic [17]. The spontaneous plumbagin-albumin adduct described here
involves the formation of a covalent, stable bond possibly between cysteine 34 of albumin and carbon
C3 of plumbagin. Our finding is consistent with the long-known reactivity of 2-methyl-1,4-
naphthoquinones with sulfhydryl groups and cysteine, which involves the 3-position on the quinone
ring [52, 53]. The thiolate group of cysteine-34 in human serum albumin is reactive under
physiological pH because of its low pKa (~6.5) [45, 54]. It is therefore exploited to conjugate drugs to
albumin as a means of delivery (reviewed in [55]). Considering the high reactivity of plumbagin
toward cysteine and cysteine-containing nucleophiles, together with the presence of a free cysteine
residue in serum albumin in physiological conditions, the formation of the plumbagin-albumin adduct
is expected. Previous reports have described the binding of naphthoquinones with cysteine in albumin
and, most importantly, their detection in vivo. For example, treatment of rats with naphthalene induced
cysteinyl adducts of both hemoglobin and albumin with naphthalene metabolites 1,2-naphthoquinone
and 1,4-naphthoquinone in a dose-dependent manner [56] whereas serum albumin adducts of both 1,2-
naphthoquinone and 1,4-naphthoquinone metabolites have been detected in the blood of human
subjects [57].
Overall, the binding of plumbagin to serum albumin naturally solves the problems of plumbagin’s poor
aqueous solubility, high lipophilicity and instability (reactivity), which were thought likely to impede
its clinical translation by limiting its bioavailability, in vivo absorption, distribution and tumor uptake.
Indeed, many laboratories, including ours [58], have designed new formulations and delivery systems
such as liposomes, niosomes, microspheres, nanoparticles, micelles, complexation, metal
nanoparticles, crystals modification, etc (reviewed in [59]). Our new findings suggest that these efforts
may be unnecessary if conjugation to albumin improves the pharmacokinetic profile of plumbagin. In
agreement with this notion, we observed that despite its rapid clearance in plasma, plumbagin
displayed long-lasting effects in vivo. A mouse prostate tumor model was used to compare several
plumbagin administration schedules as previously described [8, 9]. As shown in Supplementary
Figure S8, and consistent with previous findings, per oral plumbagin at 1mg/kg caused tumors to
shrink over time. Similar efficacy was observed when plumbagin was administered once/day, every 3
days or every 5 days, whereas administration every 7 days was less efficient at decreasing tumor size
-
17 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
than other schedules. This experiment indicates that plumbagin in vivo has prolonged effects, in
apparent contradiction with its fast clearance from plasma.
Accumulation of albumin in tumors makes it particularly appealing for oncology drug development
[
60, 61] and represents a particularly relevant property considering that plumbagin will be used as an
anti-cancer agent. There are several passive and active mechanisms that facilitate the accumulation of
albumin in solid tumors, particularly enhanced permeability and retention (EPR) of macromolecules in
tumor tissue due to leaky tumor blood vessels and a lack of lymphatic drainage [19, 20]. Furthermore,
albumin-binding proteins such as albondin (gp 60) expressed on the endothelial cell surface and
SPARC (Secreted Protein, Acidic and Rich in Cysteine) present in the tumor interstitium facilitate the
uptake and retention of albumin in the tumor interstitium [14, 22]. Albumin also accumulates in tumors
because of high metabolic turnover associated with hypoalbuminaemia in patients with advanced solid
tumors [62].
In summary, plumbagin-albumin adducts are hypothesized to naturally take advantage of the desirable
properties of human serum albumin, such as better solubility, increased accumulation in tumors,
increased plasma half-life of small molecule drugs, and reduced toxicity. Spontaneous formation of a
plumbagin-albumin adduct as described here explains the rapid clearance of plumbagin from plasma
upon administration. When added to cancer cells the conjugate displayed several cellular effects
consistent with known action of plumbagin such as cell growth inhibition, cell cycle arrest, decreased
expression of AR and PKC isoforms, increased AMPK phosphorylation, and therefore represents a
biologically relevant intermediate of plumbagin.
-
18 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
REFERENCES
1
.
S. Padhye, P. Dandawate, M. Yusufi, A. Ahmad, F.H. Sarkar. Med. Res. Rev. 2012, 32: 1131-
158.
1
2
3
4
5
.
.
.
.
P. Panichayupakaranant, M.I. Ahmad. Adv. Exp. Med. Biol. 2016, 929: 229-246.
Y. Liu, Y. Cai, C. He, M. Chen, H. Li. Am. J. Chin. Med. 2017, 45: 423-441.
S.K. Tripathi, M. Panda, B.K. Biswal. Food Chem. Toxicol. 2019, 125: 566-582.
W. Yan, T.Y. Wang, Q.M. Fan, L. Du, J.K. Xu, Z.J. Zhai, H.W. Li, T.T. Tang. Acta. Pharmacol.
Sin. 2014, 35(1): 124-134.
6
.
B.B. Hafeez, W. Zhong, J.W. Fischer, A. Mustafa, X. Shi, L. Meske, H. Hong, W. Cai, T.
Havighurst, K. Kim, A.K. Verma. Mol. Oncol. 2013, 7: 428-439.
7
8
9
.
.
.
C.G. Kang, E. Im, H.J. Lee, E.O. Lee. Bioorg. Med. Chem. Lett. 2017, 27: 1914-1918.
P. Abedinpour, V.T. Baron, A. Chrastina, J. Welsh, P. Borgström, Prostate 2013, 73: 489-99.
P. Abedinpour, V.T. Baron, A. Chrastina, G. Rondeau, J. Pelayo, J. Welsh, P. Borgström,
Prostate 2017, 77: 1550-1562.
1
1
1
0. C. Kyriakopoulos, C.J. Paller, A. Verma, K. Kader, J. Kittrelle, P. Borgström, U.N.
Vaishampayan. Journal of Clinical Oncology 2019, 37: 15_suppl, e16517-e16517.
1. M.R. Kumar, B.K. Aithal, N. Udupa, M.S. Reddy, V. Raakesh, R.S. Murthy, D.P. Raju, B.S.
Rao. Drug Deliv. 2011, 18: 511-522.
2. Y.J. Hsieh, L.C. Lin, T.H. Tsai. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 2006,
8
44(1): 1-5.
1
1
1
1
1
3. B.R. Jali, Y. Kuang, N. Neamati, J.B. Baruah. Chem. Biol. Interact. 2014, 214: 10-17.
4. F. Kratz. J. Control Release 2014, 190: 331-336.
5. E.N. Hoogenboezem, C.L. Duvall. Adv. Drug Deliv. Rev. 2018, 130: 73-89.
6. F. Kratz, B. Elsadek. J. Control Release 2012, 161(2): 429-445.
7. K. Yamasaki, V.T. Chuang, T. Maruyama, M. Otagiri. Biochim. Biophys. Acta. 2013, 1830:
5
435-5443.
1
1
2
8. D. Sleep, J. Cameron, L.R. Evans. Biochim. Biophys. Acta. 2013, 1830: 5526-5534.
9. H. Maeda, J. Wu, T. Sawa, Y. Matsumura, K. Hori. J. Control Release 2000, 65(1-2): 271-284.
0. A. Mukherjee, B. Kumar, K. Hatano, L.M. Russell, B.J. Trock, P.C. Searson, A.K. Meeker, M.G.
Pomper, S.E. Lupold. Mol. Cancer Ther. 2016, 15(10): 2541-2550.
2
2
1. J.E. Schnitzer. Am. J. Physiol. 1992, 262(1 Pt 2): H246-54.
2. A.M. Merlot, D.S. Kalinowski, D.R. Richardson. Front. Physiol. 2014, 5: 299.
-
19 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
2
2
3. J.E. Schnitzer, A. Sung, R. Horvat, J. Bravo. J. Biol. Chem. 1992, 267(34): 24544-24553.
4. J. Jiao, S. Wang, R. Qiao, I. Vivanco, P.A. Watson, C.L. Sawyers, H. Wu: Cancer Res. 2007, 67:
6
083-6091.
2
2
2
2
2
5. R. Pio, Z. Jia, V.T. Baron, D. Mercola, UCI NCI SPECS. PLoS One 2013, 8(1): e54096.
6. P.S. Guin, S. Das, P.C. Mandal. Int. J. Electrochem. 2011, 2011: 1-22.
7. N.J. Brunskill. Nephrol. Dial. Transplant. 2000, 15(11): 1732-1734.
8. E. Frei. Diabetol. Metab. Syndr. 2011, 3(1):11.
9. M. Damle, L.A.A. Newton, M. M. Villalba, R. Leslie, J. Davis. Electroanalysis 2010, 22(21):
2
491–2495.
3
0. G. Rondeau, P. Abedinpour, A. Chrastina, J. Pelayo, P. Borgstrom, J. Welsh. Sci. Rep. 2018,
8
(1): 2694.
3
3
1. M.H. Aziz, N.E. Dreckschmidt, A.K. Verma, Cancer Res. 2008, 68: 9024-9032.
2. B.B. Hafeez, W. Zhong, A. Mustafa, J.W. Fischer, O. Witkowsky, A.K. Verma. Carcinogenesis
2
012, 33: 2586-2592.
3
3
3
3. B.B. Hafeez, J.W. Fischer, A. Singh, W. Zhong, A. Mustafa, L. Meske, M.O. Sheikhani, A.K.
Verma. Cancer Prev. Res. 2015, 8: 375-386.
4. M.B. Chen,Y. Zhang, M.X. Wei, W. Shen, X.Y. Wu, C. Yao, P.H. Lu. Cell Signal. 2013, 25:
1
993-2002.
5. H. Wang, H. Zhang, Y. Zhang, D. Wang, X. Cheng, F. Yang, Q. Zhang, Z. Xue, Y. Li, L. Zhang,
L. Yang, G. Miao, D. Li, Z. Guan, Y. Da, Z. Yao, F. Gao, L. Qiao, L. Kong, R. Zhang.
Oncotarget. 2016, 7: 82864-82875.
3
3
3
6. M.T. Larsen, M. Kuhlmann, M.L. Hvam, K.A. Howard. Mol. Cell. Ther. 2016, 4: 3.
7. N. Gautam, R. Thakare, S. Rana, A. Natarajan, Y. Alnouti. Xenobiotica (2015) 45(10): 858-873.
8. J. Wang, X.X. Li-Chan, J. Atherton, L. Deng, R. Espina, L. Yu, P. Horwatt, S. Ross, S.
Lockhead, S. Ahmad, A. Chandrasekaran, A. Oganesian, J. Scatina, A. Mutlib, R. Talaat. Drug
Metab. Dispos. 2010, 38(7): 1083-1093.
3
4
4
4
9. C. Prakash, R. Sharma, M. Gleave, A. Nedderman. Curr. Drug Metab. 2008, 9(9): 952-964.
0. W. Tang, A.Y. Lu. Drug. Metab. Rev. 2010, 42(2): 225-249.
1. F. Li, J. Lu, X. Ma. Chem. Res. Toxicol. 2011, 24(5): 744-751.
2. M. Sogami, S. Era, S. Nagaoka, K. Kuwata, K. Kida, J. Shigemi, K. Miura, E. Suzuki, Y. Muto,
E. Tomita, et al. J. Chromatogr. 1985, 332: 19-27.
4
3. S. Era, T. Hamaguchi, M. Sogami, K. Kuwata, E. Suzuki, K. Miura, K. Kawai, Y. Kitazawa, H.
Okabe, A. Noma, et al. Int. J. Pept. Protein Res. 1988, 31: 435-442.
-
20 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
4
4
4. T. Etoh, M. Miyazaki, K. Harada, M. Nakayama, A. Sugii. J. Chromatogr. 1992, 578: 292-296.
5. A.J. Stewart, C.A. Blindauer, S. Berezenko, D. Sleep, D. Tooth, P.J. Sadler. FEBS J. 2005,
2
72(2): 353-362.
4
4
6. L. Turell, R. Radi, B. Alvarez. Free Radic. Biol. Med. 2013, 65: 244-253.
7. Y. Kumagai, Y. Shinkai, T. Miura, A.K. Cho. Annu. Rev. Pharmacol. Toxicol. 2012, 52: 221-
2
47.
4
4
5
8. A.S. Ali, S. Ali, B.F. El-Rayes, P.A. Philip, F.H. Sarkar. Cancer Treat. Rev. 2009, 35: 1-8.
9. E. Totoń, E. Ignatowicz, K. Skrzeczkowska, M. Rybczyńska. Pharmacol. Rep. 2011, 63: 19-29.
0. A. Gutierrez-Uzquiza, C. Lopez-Haber, D.L. Jernigan, A. Fatatis, M.G. Kazanietz. Mol. Cancer
Res. 2015 13: 1336-1346.
5
5
5
5
5
1. S.M. Jeon. Exp. Mol. Med. 2016, 48: e245.
2. C.A. Colwell, M. McCall. Science 1945, 101(2632): 592-594.
3. L.F. Fieser, R.B. Turner. J. Am. Chem. Soc. 1947, 69(10): 2335-2338.
4. A.O. Pedersen, J. Jacobsen. Eur. J. Biochem. 1980, 106: 291-295.
5. M.B. Caspersen, M. Kuhlmann, K. Nicholls, M.J. Saxton, B. Andersen, K. Bunting, J. Cameron,
K.A. Howard. Ther. Deliv. 2017, 8(7): 511-519.
5
5
5
5
6
6. S. Waidyanatha, M.A. Troester, A.B. Lindstrom, S.M. Rappaport. Chem. Biol. Interact. 2002,
1
41(3): 189-210.
7. P.H. Lin, D.R. Chen, T.W. Wang, C.H. Lin, M.C. Chuang. Chem. Biol. Interact. 2009, 181(1):
1
07-114.
8. A. Chrastina, V.T. Baron, P. Abedinpour, G. Rondeau, J. Welsh, P. Borgström. Biomed. Res. Int.
2
018, 2018: 9035452.
9. S. Rajalakshmi, N. Vyawahare, A. Pawar, P. Mahaparale, B. Chellampillai. Artif. Cells
Nanomed. Biotechnol. 2018, 46(sup1): 209-218.
0. G. Stehle, H. Sinn, A. Wunder, H.H. Schrenk, S. Schütt, W. Maier-Borst, D.L. Heene.
Anticancer Drugs. 1997, 8: 677-685.
6
6
1. P. Kremer, M. Fardanesh, R. Ding, M. Pritsch, S. Zoubaa, E. Frei. Neurosurgery 2009, 64: 53-60.
2. C.A. Hauser, M.R. Stockler, M.H. Tattersall. Support. Care Cancer. 2006, 14(10): 999-1011.
-
21 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
LEGENDS
Figure 1. Spectral analysis of complex formation.
Panel A: structure of plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone). Panel B: structure of
DMNQ (2,3-dimethoxy-1,4-naphthoquinone). Panel C: Reference light absorption spectra of 100
μmol/L plumbagin, DMNQ and 4.5 % BSA (w/v) in PBS at pH7.4. The peak absorption of plumbagin
was at 418 nm and the peak absorption of DMNQ was at 339 nm. Panel D: Spectral changes of 100
μmol/L plumbagin (left) and 100 μmol/L DMNQ (right) mixed with albumin were monitored at
various time points (30 sec - 6h) starting immediately upon incubation with BSA (45 mg/ml) in PBS at
pH7.4. UV-VIS light absorption spectra were acquired using a DU-640 spectrophotometer (Backman
Coulter). Insets show the progressive decrease in absorbance at the peak absorption of plumbagin (418
nm), a change not observed at the peak absorption of DMNQ (339 nm).
Figure 2: Changes in plumbagin redox activity upon complex formation
Panel A: Concentration-dependent cyclic voltammograms of plumbagin (concentration range 0.2-100
μmol/L) in PBS at pH 7.4. Working electrode, GC, counter electrode, Pt, reference electrode,
Ag/AgCl/sat, scan rate, v= 100 mV/s. Panel B: Peak anodic current (Ipa) as a function of plumbagin
concentration. Panel C: Time-dependent changes in the redox activity of plumbagin were measured by
recording cyclic voltammograms at 10 sec, 30 min, 1 and 3 h after addition of 100 μmol/L plumbagin
to BSA (45 mg/ml) in PBS at pH 7.4. The arrows indicate the direction of time-dependent changes
from 10 sec to 3 h. Panel D: Scaled-up CVs of plumbagin. Panel E: Time-dependent changes in the
redox activity of 100 μmol/L DMNQ mixed with BSA in the same conditions as plumbagin. Panel F:
Scaled-up CVs of DMNQ. DMNQ did not show further variation in I-E profile and in redox peak
currents. Experimental conditions: working electrode, GC, counter electrode, Pt, reference electrode,
Ag/AgCl/sat, scan rate, v= 100 mV/s. Panel G: Changes in peak anodic current (ipa) over time for
plumbagin and DMNQ in BSA. The ipa values were determined from cyclic voltammograms of 100
μmol/L plumbagin or DMNQ incubated with BSA, acquired at indicated time points. The dotted line
(.....) shows the level of capacitive current (unconjugated BSA in PBS).
Figure 3: Extraction recovery of plumbagin and 1,4-naphthoquinone analogs to albumin.
Panel A: Warfarin was added to BSA (45 mg/ml in PBS, pH7.4, 37°C) at a final concentration of 100
µmol/L. The percentage of free warfarin and of warfarin bound to BSA as estimated by membrane
ultrafiltration using Amicon Ultracell 10 K (Millipore). Panel B: Extraction recovery of warfarin from
-
22 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
albumin. Non-covalently bound warfarin was extracted by chloroform at the indicated time points.
Amount of extracted drug was determined by UV-VIS spectroscopy as described in methods. The ratio
of recovered (free) drug from the complex was calculated at the indicated time points. Panel C:
Extraction recovery of plumbagin from albumin. Plumbagin was added to BSA (45 mg/ml in PBS,
pH7.4, 37°C for 18h) at final concentrations of 10 and 100 μmol/L. Panel D: Structure of 1,4-
naphthoquinones structurally related to plumbagin. Plumbagin (A), juglone (B), naphthazarin (C), 1,4-
naphthoquinone (D), menadione (E), 5-Oacetyl-plumbagin (F), 2,3-dimethoxy-1,4-naphthoquinone (G)
and 2,3-dimethyl-1,4-naphthoquinone (H). Panel E: Recovery of 1,4-naphthoquinones from BSA.
Each 1,4-naphthoquinone (100 μmol/L) was incubated with BSA (45 mg/ml in PBS, pH 7.4, 37°C for
1
8h). Recovery indicates reversible non-covalent interaction with serum albumin, while lack of
recovery demonstrates adduct formation.
Figure 4: Species specificity and lysosomal-dependency of the effect of plumbagin-albumin
Panel A: Mouse and human cancer cells were plated 24h before the experiment. Plumbagin-albumin
was added at increasing concentrations up to 20 µmol/L of plumbagin for 24h. Unconjugated BSA or
HSA were used as control as appropriate (mass equivalent). Cell numbers were evaluated using the
WST-1 assay. Each graph represents the means ±SEM of at least 3 independent experiments, each
performed in triplicates, except for UM-UC-3 (2 independent experiment, each performed in
triplicates). Panel B: PTEN-P2 cells were pre-incubated with or without various inhibitors for 3hrs
before addition of conjugate plumbagin-BSA (10 µmol/L plumbagin) or unconjugated BSA (equimolar
to albumin adduct) for 24h. Inhibitors were: serine protease inhibitor leupeptin (125 µmol/L); caspase
inhibitor Z-VAD (25 µmol/L); cathepsin D/E inhibitor pepstatin A (2.5 µmol/L); anti-oxidant Trolox
(20 µmol/L). Cell numbers were evaluated using the WST-1 assay. The graph represents the means
±
SEM of 3 independent experiments (n=2 for Trolox), each performed in triplicates. Results are
expressed as % of control (BSA alone).
Figure 5: Effect of plumbagin-albumin on cell growth
Panel A: Growth over time of cells treated with plumbagin-albumin. Conjugated plumbagin-BSA (10
µmol/L) or BSA (mass equivalent) were added to PTEN-P2 cells at day 0 for a one-time treatment
(acute treatment), or every day from day 0 (chronic treatment). The number of cells was evaluated
-
23 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
every day by measuring mitochondrial activity using the WST-1 assay. The graph represents the means
SEM of 3 separate experiments, each performed in triplicates. Results are expressed as percent of
±
maximum as a function of time. Panels B: Cell cycle analysis. PTEN-P2 cells were incubated with
BSA or with plumbagin-BSA for 24h, 48 and 72h. Cells were suspended, fixed and stained with
propidium iodide followed by flow cytometry. Cell cycle phases were analyzed using De Novo™ FCS
express. Panel C: The proportion of cells in each phase of the cell cycle was represented in pie charts.
Panel D: Western blot analysis of RBBP expression following plumbagin-albumin treatment. Cells
were incubated with increasing concentrations of plumbagin-BSA. After 24h, cells were lysed and the
relative protein expression was analyzed by western blot. The membrane was stripped and probed
again with the indicated antibodies. The insert shows the quantification of RBBP expression in this
experiment.
Figure 6: Western blot analysis of AR expression following plumbagin-albumin treatment.
Cells were incubated with plumbagin-BSA or with unconjugated BSA for the indicated times. Cells
were lysed and relative protein amounts were analyzed by western blot. The membrane was stripped
and probed several times with the indicated antibodies. Panel A: PTEN-P2 cells were treated with
increasing concentrations of plumbagin-BSA for 24h. The insert shows the quantification of AR
expression in this experiment. Panel B: PTEN-P2 cells were treated with 10 µmol/L plumbagin-BSA
for the indicated times. Panel C: Average AR expression following treatment with 10 µmol/L of
plumbagin conjugate for 24h. Values represent means ± SE of 5 separate experiments. Panel D:
Human cancer cells LNCaP were treated with plumbagin-HSA at 10 µmol/L for 24h.
Figure 7: Western blot analysis of PKC expression and AMPK phosphorylation following
plumbagin-albumin treatment.
Cells were incubated with plumbagin-BSA or with BSA for various times of treatment, lysed, and the
relative protein expression was analyzed by western blot. The membrane was stripped and probed
several times. Panel A: PTEN-P2 cells were treated with 10 µmol/L plumbagin-BSA for up to 48h.
PKCε, PKCµ and PKCδ were sequentially analyzed by western blot using isoform-specific antibodies.
Panel B: Average PKCε expression following treatment with 10µM of plumbagin conjugate for 24h.
Values represent means ± SE of 5 separate experiments. Panel C: PTEN-P2 cells were treated with
-
24 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
1
0µM plumbagin-BSA for up to 4 hours. A pan phospho-PKC (βII-Ser660) antibody was used to
measure PKC phosphorylation. Phospho-AMPKα (Thr172) and phospho-ACC (Ser79) were
sequentially analyzed by western blot using phosphor-isoform-specific antibodies. Panel D: Average
AMPKα phosphorylation following treatment with 10 µmol/L of plumbagin conjugate for 30min and
1
h (peak). Values represent means ± SE of 5 separate measurements from 2 independent experiments.
-
25 -
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
A
B
Plumbagin
DMNQ
C
1.2
1.0
0.8
0.6
0.4
BSA
Plb
DMNQ
0.2
0.0
3
00 350 400
450 500
wavelength (nm)
D
0
0
0
0
.45
0.6
.4
0.2
.0
1
1
0
0
0
0
0
.2
.0
.8
.6
.4
.2
.0
1.2
.35
.25
.15
0
1.0
0.8
0.6
0.4
0.05
0
0
100 200 300 400
time (min)
0
100 200 300 400
time (min)
0.2
0.0
3
00
350
400
450
500
300
350
400
450
500
wavelength (nm)
wavelength (nm)
FIGURE 1
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
A
B
7
6
5
4
3
2
1
.E-06
.E-06
.E-06
.E-06
.E-06
.E-06
.E-06
0
.E+00
0
50
100
[
plumbagin]/µmol.l-1
E [V] vs Ag/AgCl
C
D
0
1
3
1
3
0sec
0min
h
t
h
t
E[V] vs Ag/AgCl
E[V] vs Ag/AgCl
E
F
t
E[V] vs Ag/AgCl
E[V] vs Ag/AgCl
G
1.6E-06
1.4E-06
1.2E-06
1.0E-06
8.0E-07
6.0E-07
4.0E-07
2.0E-07
DMNQ
Plb
0E+00
0
1
2
3
time (h)
FIGURE 2
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
A
C
120
00
B
120
100
80
60
40
20
0
1
80
60
40
20
0
%
free
% bound
0
5
10
Time (h)
15
20
0
5
10
Time (h)
15
20
100
8
6
4
2
0
0
0
0
0
1
0 µM
100 µM
0
5
10
Time (h)
15
c
20
D
E
a
b
d
e
f
g
h
FIGURE 3
This article is protected by copyright. All rights reserved.

ChemMedChem
10.1002/cmdc.202000157
A
B
-
Leu Z pepA Trlx
BSA
-
Leu Z pepA Trlx
Plb-BSA
FIGURE 4
This article is protected by copyright. All rights reserved.